Electroluminescent (EL) devices are a promising technology for flat-panel displays. For example, Organic Light Emitting Diodes (OLEDs) have been known for some years and have been recently used in commercial display devices. EL devices use thin-film layers of materials coated upon a substrate that emit light when electric current is passed through them. In OLED devices, one or more of those layers includes organic material. Using active-matrix control schemes, a plurality of EL light-emitting devices can be assembled into an EL display. EL subpixels, each including an EL device and a drive circuit, are typically arranged in two-dimensional arrays with a row and a column address for each subpixel, and are driven by a data value associated with each subpixel to emit light at a brightness corresponding to the associated data value. To make a full-color display, one or more subpixels of different colors are grouped together to form a pixel. Thus each pixel on an EL display includes one or more subpixels, e.g. red, green, and blue. The collection of all the subpixels of a particular color is commonly called a “color plane.” A monochrome display can be considered to be a special case of a color display having only one color plane.
Typical large-format displays (e.g. having a diagonal of greater than 12 to 20 inches) employ hydrogenated amorphous silicon thin-film transistors (a-Si TFTs) formed on a substrate to drive the subpixels in such large-format displays. Amorphous Si backplanes are inexpensive and easy to manufacture. However, as described in “Threshold Voltage Instability Of Amorphous Silicon Thin-Film Transistors Under Constant Current Stress” by Jahinuzzaman et al. in Applied Physics Letters 87, 023502 (2005), the a-Si TFTs exhibit a metastable shift in threshold voltage (Vth) when subjected to prolonged gate bias. This shift is not significant in traditional display devices such as LCDs, because the current required to switch the liquid crystals in LCD display is relatively small. However, for LED applications, much larger currents must be switched by the a-Si TFT circuits to drive the EL materials to emit light. Thus, EL displays employing a-Si TFT circuits generally exhibit a significant Vth shift as they are used. This Vth shift can result in decreased dynamic range and image artifacts. Moreover, the organic materials in OLED and hybrid EL devices also deteriorate in relation to the integrated current density passed through them over time, so that their efficiency drops while their resistance to current, and thus forward voltage, increases. These effects are described in the art as “aging” effects.
These two factors, TFT and EL aging, reduce the lifetime of the display. Different organic materials on a display can age at different rates, causing differential color aging and a display whose white point varies as the display is used. If some EL devices in the display are used more than others, spatially differentiated aging can result, causing portions of the display to be dimmer than other portions when driven with a similar signal. This can result in visible burn-in. For example, this occurs when the screen displays a single graphic element in one location for a long period time. Such graphic elements can include stripes or rectangles with background information, e.g. news headlines, sports scores, and network logos. Differences in signal format are also problematic. For example, displaying a widescreen (16:9 aspect ratio) image letterboxed on a conventional screen (4:3 aspect ratio) requires the display to matte the image, causing the 16:9 image to appear on a middle horizontal region of the display screen and black (non-illuminated) bars to appear on the respective top and bottom horizontal regions of the 4:3 display screen. This produces sharp transitions between the 16:9 image area and the non-illuminated (matte) areas. These transitions can burn in over time and become visible as horizontal edges. Furthermore, the matte areas are not aged as quickly as the image area in these cases, which can result in the matte areas' being objectionably brighter than the 16:9 image area when a 4:3 (full-screen) image is displayed.
One approach to avoiding the problem of voltage threshold shift in TFT circuits is to employ circuit designs whose performance is relatively constant in the presence of such voltage shifts. For example, U.S. Patent Application Publication No. 2005/0269959 filed by Uchino et al, Dec. 8, 2005, entitled “Pixel Circuit, Active Matrix Apparatus And Display Apparatus” describes a subpixel circuit having a function of compensating for characteristic variation of an electro-optical element and threshold voltage variation of a transistor. The subpixel circuit includes an electro-optical element, a holding capacitor, and five-channel thin-film transistors. Alternative circuit designs employ current-mirror driving circuits that reduce susceptibility to transistor performance. For example, U.S. Patent Application Publication No. 2005/0180083 filed by Takahara et al., Aug. 15, 2005 entitled “Drive Circuit For EL Display Panel” describes such a circuit. However, such circuits are typically much larger and more complex than the two-transistor, single capacitor (2T1C) circuits otherwise employed, thereby reducing the aperture ratio (AR), the percent of the area on a display available for emitting light. The decrease in AR decreases the display lifetime by increasing the current density through each EL device.
Other methods used with a-Si TFTs rely upon measuring the threshold-voltage shift. For example, U.S. Patent Application Publication No. 2004/0100430A1 “Active Matrix Drive Circuit” by Fruehauf, published May 27, 2004, describes an OLED subpixel circuit including a conventional 2T1C subpixel circuit and a third transistor used to carry a current to an off-panel current measurement circuit. As Vth shifts and the OLED ages, the current decreases. This decrease in current is measured and used to adjust the data value used to drive the subpixel. Similarly, U.S. Pat. No. 6,433,488 B1 “OLED Active Driving System with Current Feedback” by Bu, granted Aug. 13, 2002, describes using a third transistor to measure the current flowing through an OLED device under a test condition and comparing that current to a reference current to adjust the data value. Additionally, Arnold et al., in commonly-assigned U.S. Pat. No. 6,995,519, granted Feb. 7, 2006, teach using a third transistor to produce a feedback signal representing the voltage across the OLED, allowing for compensation of OLED aging but not Vth shift. However, although these schemes do not require as many transistors as subpixel circuits with internal compensation, they do require additional signal lines on a display backplane to carry the measurements. These additional signal lines reduce aperture ratio and add assembly cost. For example, these schemes can require one additional data line per column. This doubles the number of lines that have to be bonded to driver integrated circuits, increasing the cost of an assembled display, and increasing the probability of bond failure, thus decreasing the yield of good displays from the assembly line. This problem is particularly acute for large-format, high-resolution displays, which can have over two thousand columns. However, it also affects smaller displays, as higher bondout counts can require higher-density connections, which are more expensive to manufacture and have lower yield than lower-density connections.
Alternative schemes for reducing image burn-in have been addressed for televisions using a cathode ray tube display. U.S. Pat. No. 6,359,398 entitled “Method to Control CRT Phosphor Aging” issued Mar. 19, 2002, describes methods and apparatus that are provided for equally aging a cathode ray tube (CRT). Under this scheme, when displaying an image of one aspect ration on a display of a different aspect ratio, the matte areas of the display are driven with an equalization video signal. In this manner, the CRT is uniformly aged. However, the solution proposed requires the use of a blocking structure such as doors or covers that can be manually or automatically provided to shield the matte areas from view when the equalization video signal is applied to the otherwise non-illuminated region of the display. This solution is unlikely to be acceptable to most viewers because of the cost and inconvenience. U.S. Pat. No. 6,359,398 also discloses that matte areas can be illuminated with gray video having luminance intensity matched to an estimate of the average luminous intensity of the program video displayed in the primary region. As indicated therein, however, such estimation is not perfect, resulting in a reduced, but still present, non-uniform aging.
U.S. Pat. No. 6,369,851 entitled “Method and Apparatus to Minimize Burn Lines in a Display” issued Apr. 9, 2002 describes a method and apparatus for displaying a video signal using an edge modification signal to reduce spatial frequency and minimize edge burn lines, or a border modification signal to increase brightness of image content in a border area of a displayed image, where the border area corresponds to a non-image area when displaying images with a different aspect ratio. However, these solutions can cause objectionable image artifacts, for example reduced sharpness or visibly brighter border areas in displayed images.
The general problem of regional brightness differences due to burn-in of specific areas due to video content has been addressed in the prior art, for example by U.S. Pat. No. 6,856,328 entitled, “System and method of displaying images.” This disclosure teaches that the burn-in of graphic elements as described above can be prevented by detecting those elements in the corners of the image and reducing their intensity to the average display load. This method requires the detection of static areas and may not prevent color-differentiated burn-in. An alternative technique is described in Japanese Publication No. 2005-037843 A by Igarashi et al. entitled “Camera and Display Control Device”. In this disclosure, a digital camera is provided with an organic EL display that is prevented from burning in by employing a DSP in the digital camera. The DSP changes the position of an icon on the organic EL display by changing the position of the icon image data in a memory every time that the camera is turned on. Since the degree to which the display position is changed is approximately one pixel a user cannot recognize the change in the display position. However, this approach requires a prior knowledge and control of the image signal and does not address the problem of format differences.
U.S. Patent Application Publication No. 2005/0204313 A1 by Enoki et al. describes a further method for display screen burn prevention, wherein an image is gradually moved in an oblique direction in a specified display mode. This and similar techniques are generally called “pixel orbiter” techniques. Enoki et al. teach moving the image as long as it displays a still image, or at predetermined intervals. Kota et al., in U.S. Pat. No. 7,038,668, granted May 2, 2006, teach displaying the image in a different position for each of a predetermined number of frames. Similarly, commercial plasma television products advertise pixel orbiter operational modes that sequentially shift the image three pixels in four directions according to a user-adjustable timer. However, these techniques may not employ all pixels of a display, and therefore may create a border effect of pixels that are brighter than those pixels in the image area that are always used to display image data.
Existing methods for mitigating image burn-in on EL displays generally either require additional display circuitry or manipulate the displayed image. Methods requiring additional display circuitry can reduce the lifetime of the display, increase its cost, and reduce manufacturing yield. Methods manipulating the displayed image cannot correct for all burn-in. Accordingly, there is a need for an improved method and apparatus for providing improved display uniformity in electroluminescent flat-panel display devices.